THE EFFECT OF AGING ON THE CLUSTERING AND PRECIPITATION PROCESS IN Al-Mg-Si ALLOYS

The dependence of the mechanism of age hardening in Al-Mg-Si alloys on their Si content, the pre-aging conditions and addition of Cu has been investigated. For this purpose, the solute clusters and the metastable precipitates in aged Al-Mg-Si alloys have been characterized by a three dimensional atom probe (3DAP) and transmission electron microscopy (TEM). Atom probe analysis results have revealed that the Mg:Si atomic ratio in the co-clusters, GP zones and β′′ is close to that of the alloy composition. This finding suggests that excess Si causes a higher density of clusters and/or precipitates and leads to pronounced age hardening response. On the other hand, the density of GP zones and β′′ after artificial aging at 175°C depends on pre-aging conditions, for example, pre-aging at 70°C increase the density of GP zones and β′′, whereas natural aging reduces it. Based on these results, the characteristic two-step age-hardening response and the precipitation process of Al-Mg-Si(-Cu) alloys are discussed. Introduction Numerous investigations have been carried out on the precipitation process [1-3], the mechanism of two-step aging [4] and the effect of alloy composition on the age-hardening response [5,6] of Al-Mg-Si(-Cu) alloys. Since Al-Mg-Si(-Cu) alloys are considered to be the most promising candidates for heat-treatable bodysheet materials, study of the precipitation process in these alloys has been received considerable attention by the automobile industry. In the automobile manufacturing process, substantial age-hardening must occur during artificial aging in less than 30 min at about 170°C, because the alloys for body sheet applications must be age-hardened during the paint-baking process. It was reported that the Al-Mg-Si alloys containing an excess amount of Si than the Al-Mg2Si quasibinary composition exhibit more rapid age hardening response [5]. On the other hand, the hardening response is significantly suppressed when the alloy receives a room temperature aging for a prolonged period of time. Since natural aging can not be avoided in the automobile manufacturing process, understanding of the mechanism of this adverse age hardening effect is strongly desired. The first direct observation of the formation of solute clusters was reported by Edwards et al. [7-9] Using the atom probe field ion microscopy (APFIM), the formation of not only the separate Siclusters but also separate Mgclusters and Mg-Si co-clusters during 70°C aging was reported. However, the solute clustering during natural aging was not investigated in their study. Furthermore, the influence of the alloy composition is not clearly understood yet. Characterizing features of possible solute clusters during natural aging is important in order to understand the mechanism of adverse age hardening effect due to natural aging and the role of excess amount of Si on changing the age hardening response. There are several reports on the effect of quaternary element additions on age-hardening of Al-Mg-Si alloys [1,10-12]. Cu addition, in general increases the kinetics of precipitation during artificial aging. It also gives beneficial effect in reducing the deterioration of the age-hardening response arising from natural aging of Al-Mg-Si alloy. Recently, Laughlin et al. [12] reported that the Cu level has a large effect on the hardening kinetics especially in the underaged regime and a smaller but noticeable effect on the maximum hardness. Since the typical paint-bake cycle in the automobile manufacturing process involves 30 min heating at around 175°C, the body sheet aluminum alloys must be used in an underaged condition. Therefore, it is important to investigate the effect of Cu in the early stage of artificial aging. The present study aimed at understanding the dependence of the mechanism of age hardening of Al-Mg-Si alloys is on their Si content and the pre-aging conditions. The role of Cu addition to an Al-Mg-Si alloy in enhancing the age hardening response was also investgated. For this purpose, the chemical compositions of the precipitation products in Al-Mg-Si(-Cu) alloys after artificial aging at 175°C has been identified using a three dimensional atom probe (3DAP) and transmission electron microscopy (TEM). 3DAP is capable of mapping individual atoms in the real space with a near-atomic resolution [13,14], thus it can determine chemical compositions of nanoscale precipitates embedded in a matrix phase without any convolution effect. Experimental Chemical compositions of the alloys used in this study are Al-0.70Mg-0.33Si (balance), Al-0.65Mg-0.70Si (Si-excess) and Al-0.61Mg-1.22Si-0.39Cu (Cu bearing) (all in at. %). These alloys were solution treated at 550°C or 525°C for 30 min and subsequently water quenched. The solution treated specimens were subjected to various heat treatments including natural aging for 70 days, pre-aging for 16 h at 70°C, artificial aging for 10 h at 175°C and artificial aging after the pre-aging (two-step aging). For atom probe analyses, an energy compensated time-of-flight one dimensional atom probe (1DAP) and a three dimensional atom probe (3DAP) equipped with CAMECA's tomographic atom probe (TAP) detection system [14] were used. Atom probe analyses were carried out at about 30 K with a pulse fraction (Vp/Vdc) of 20% in UHV (~1x10 Torr). Microstructures of the specimens were examined with a transmission electron microscope (TEM), Philips CM200, operated at 200kV. High resolution electron microscopy observations were carried out using JEOL JEM-2000EX, operated at 200 kV. Results and Discussion An HREM image of a naturally aged Si-excess alloy shows a uniform fringe contrast as shown in Fig. 1 (a). No contrast which can be attributed to the precipitates is observed. On the other hand, the contrast arising from the precipitates is observed in the Si-excess alloy that was pre-aged at 70°C for 16 h (Fig. 1 (b)). The HREM image indicates that the precipitates are approximately 2 nm and are coherent with the matrix, justifying that the designation as GP zones is appropriate. Figure 2 (a) shows integral profiles of Si and Mg atoms or ladder plots of the naturally aged Si-excess alloy, where the number of detected solute atoms is plotted as a function of the total number of detected atoms. The slopes of the plots represent the local concentration of the alloy, and the horizontal axis corresponds to the depth. Steep changes in the slope are recognized (indicated by arrowheads) in both Mg and Si ladder diagrams. In these regions, the concentration of Mg or Si is significantly higher than the average concentration in the alloy, suggesting that there are separate clusters of Mg and Si atoms (indicated by the arrowheads). In addition, a cocluster of Mg and Si atoms is detected as indicated by the broken lines in Fig. 2 (a). The ratio of the number of Mg and Si atoms in this co-cluster is close to 1. From these results, it can be concluded that Mg-Si co-clusters are present in the naturally aged specimen. Figure 2 (b) shows 3DAP elemental maps of Mg and Si atoms obtained from the Siexcess alloy after 70°C pre-aging. Note that the size of the dots 2nm (a)